Principles of Drug Action I, Winter 2005

1
ISOMERISM AND STEREOCHEMISTRY
Jack DeRuiter
I. Basic Principles
Isomers are defined as molecules of identical atomic compositions (molecular formulas), but
with different bonding arrangements of atoms or orientation of their atoms in space. Based on
this definition, several types of isomerism are possible including constitutional, configurational,
and conformational isomerism. Constitutional isomers (also called structural or positional
isomers) are molecules with the same atomic composition but different bonding arrangements
between atoms, as illustrated by the examples of catechol, resorcinol, and hydroquinone (Fig. 1).
All of these compounds have the same atomic composition (C
6
H
6
0
2
)
,
but different bonding
arrangements of atoms and are thus distinct chemical entities with different chemical and
physical properties.
Figure 1. Constitutional isomers
Configurational isomers are defined as molecules of identical atomic composition and bonding
arrangements of atoms, but different orientations of atoms in space, and these different
orientations cannot interconvert freely by bond rotation. Since these types of isomers differ only
in relative spatial orientations of atoms, they are commonly referred to as stereoisomers.
Configurational stereoisomers are subcategorized as optical isomers (enantiomers) or geometric
isomers (Fig. 2), depending upon the hybridization state and geometry of the atoms that impart
the properties of stereoisomerism and the overall structure of the molecule. Stereoisomers of this
type are distinct chemical entities that may have different chemical and physical 0roperties.
Conformational isomers (conformers) are stereoisomeric forms characterized by different relative
spatial arrangements of atoms that result from rotation about sigma bonds. Thus, unlike
configurational isomers, conformers are interconverting stereochemical forms of a single
compound. The nature of conformational and configurational stereoisomerism, as well as the
role of stereoisomerism in drug activity is the subject of this article.
OH
OH
OH
OH
OH
OH
Catechol
Resorcinol Hydroquinone
Principles of Drug Action I, Winter 2005
2
Figure 2. Stereoisomers
II. Optical Activity and Molecular Structure
Modern stereochemistry originated with the research of Malus in 1808 who discovered that
plane-polarized light is generated when a beam of light is passed through calcium carbonate. In
1813, the mineralogist Biot reported that asymmetrically cut quartz crystals rotate the plane of a
beam of polarized light. It also was noted that certain organic liquids, as well as solutions of
certain organic compounds, can rotate the plane of polarized light. Biot attributed this effect on
plane-polarized light to a property of the individual organic molecules through which the light is
passed, a property now referred to as optical activity. The concept of optical activity was
extended by Herschel in 1812 who observed that hemihedral quartz crystals, having odd faces
inclined in one direction, rotated the plane of polarized light in one direction, whereas crystals
whose odd faces were inclined in the opposite direction rotated plane-polarized light to the same
extent but in the opposite direction.
Pasteur refined the observations of the mineralogists by proposing a link between optical activity
and molecular structure. His landmark work of 1847 was based on earlier observations by Biot
that chemically identical salts of tartaric acid rotated plane-polarized light differently. Pasteur
discovered that two distinct crystalline forms of tartaric acid salt could be obtained from
solutions of the optically inactive salt of "paratartaric acid" (also known as racemic acid), and
that one crystal form has hemihedral faces that inclined to the right, whereas the other has faces
that inclined to the left. He separated the distinct crystalline salts forms and observed that they,
unlike paratartaric acid, are optically active; solutions of the left-handed crystals rotate the plane
of polarized light to the right, and solutions of the right-handed crystals rotate the light to the
same degree, but in the opposite direction. Pasteur further demonstrated that the left- and right-
handed crystals were mirror images of each other and concluded that this property must reflect
the handedness of the molecules that constitute the crystals.
The molecular basis for the left- and right-handedness of distinct crystals of the same chemical
substance and the associated differences in optical rotation was developed from the hypothesis of
Paterno (1869) and Kekule that the geometry about a carbon atom bound to four ligands is
tetrahedral. Based on the concept of tetrahedral geometry, Van't Hoff and LeBel concluded that
when four different groups or atoms are bound to a carbon atom, two distinct tetrahedral
molecular forms are possible, and these bear a nonsuperimposable mirror-image relationship to
one another (Fig. 3). This hypothesis provided the link between three-dimensional molecular
structure and optical activity, and represents the foundation of stereoisomerism.
W
C
Z X
Y
W
C
Z X
Y
Optical Isomers (Enantiomers)
W
C
Z
C
X
Y
W
C
Z
C
Y
X
Geometric Isomers
Principles of Drug Action I, Winter 2005
3
Figure 3. Tetrahedral geometry and optical isomerism
III. Chirality and Optical Isomers (Enantiomers)
The property of nonsuperimposability became known as chirality, and molecules containing
asymmetrically substituted carbons are referred to as chiral molecules. The term chiral was
derived from the Greek word meaning "hand" and was applied as a description of the left- and
right-handedness of crystal structure resulting from molecular asymmetry. The individual mirror
image forms of a chiral molecule are called optical isomers because they rotate the plane of
polarized light (are optically active) and differ in structure only in the orientation of atoms or
groups about the asymmetric carbon (are isomers). Today, optical isomers are more commonly
referred to as enantiomers or an enantiomeric pair.
Generally, optical isomers or enantiomers have identical physical and chemical properties; for
example, the enantiomeric forms of amphetamine (Fig. 4) have identical melting points, pKa,
solubilities, etc. There are, however, two important differences in properties between the
members of an enantiomeric pair. First, each member rotates the plane of polarized light to the
same degree, but in opposite directions. The enantiomer rotating the plane to right (clockwise) is
designated as the dextrorotatory (d) or ( +)-enantiomer. The other enantiomer rotates the plane to
the left (counterclockwise) and is designated as the levorotatory (1) or (-)-enantiomer. This is
illustrated in Fig. 4 for the enantiomers of amphetamine, where
Figure 4. Amphetamine enantiomers
the enantiomer with the specific optical rotation of (+)-21.8

is designated as dextrorotatory,
whereas the mirror enantiomer with a specific rotation of (-)-21.8

is called levorotatory. A
second difference between enantiomers is their interactions with other chiral substances. For
example, enantiomers may have different solubilities in chiral solvents, they may react at
different rates in the presence of an optically active reagent or enzyme, and many have different
affinities for chiral surfaces and receptors.
Most optically active drugs are chiral as a result of the presence of an asymmetrically substituted
W
C
Z X
Y
W
C
Z X
Y
H CH
3
NH
2
H H
3
C
NH
2
(+)-Amphetamine
[] = +21.8
o
[] = -21.8
o
(-)-Amphetamine

Principles of Drug Action I, Winter 2005
4
tetrahedral carbon atom. However, chirality can result from other types of molecular asymmetry.
For example, the presence of any asymmetrically substituted atom of tetrahedral geometry, such
as silicon, quaternary nitrogen, and metals that form tetrahedral coordination complexes
(manganese, copper, zinc, etc.) results in chirality. Similarly, compounds containing
asymmetrically substituted tetrahedral phosphorus atoms are also chiral. The antineoplastic
agent cyclophosphamide is one example of a compound with a chiral phosphorus moiety (Fig.
5).
Figure 5. Optical isomers of cyclophosphamide
Chirality is also a property of compounds containing an asymmetrically substituted atom of
pyramidal geometry. For example, secondary and tertiary amines bearing four different
substituents (one "substituent" being the lone pair of electrons on nitrogen) are chiral. However,
due to rapid pyramidal inversion (Fig. 6), the individual enantiomers are not usually separable
(resolvable). On the other hand, amines that contain an asymmetrically substituted nitrogen atom
in a ring system, particularly at a bridgehead position, may not undergo facile inversion, and thus
it is possible to resolve distinct stereoisomers (Fig. 7). The same is true for asymmetrically
substituted pyramidal sulfur derivatives such as sulfoxides, sulfonic esters, sulfonium salts, and
sulfites. For example, the sulfur atom of the nonsteroidal antiinflammatory sulindac (Fig. 8)
bears four difference substituents (one being a pair of electrons) and hence is chiral.
Figure 6. Pyramidal inversion of chiral nitrogen
Cyclic or multicyclic compounds with four different substituents projecting from the corners of
the cycle have a center of asymmetry and are chiral if all four substituents are different. Such is
the case for adamantine analogues, as shown in Fig. 9.
Certain biphenyl compounds may be chiral as a result of hindered rotation about the central bond
and dissymmetric ring substitution. In this instance bond rotation may be inhibited by the
presence of bulky groups at the ortho positions, forcing the two aromatic rings to lie in
perpendicular, or near perpendicular, planes. If each aromatic ring of such a biphenyl system is
dissymmetrically substituted, chirality results. This is illustrated by the three biphenyl
derivatives (A, B, and C) in Fig. 10. In each of these compounds rotation is restricted as a result
of the
O
P
N
O
N
Cl
Cl
H
O
P
N
O
N
Cl
Cl
H
CH
3
N
H CH
2
CH
3
..
..
CH
3
N
H CH
2
CH
3
Principles of Drug Action I, Winter 2005
5
Figure 7. Troger's base
Figure 8. The chiral sulfoxide Sulindac
Figure 9. Chiral adamantine
Figure 10. Biphenyl systems
presence of bulky ortho groups. However, only compound C is chiral because only this
compound has two dissymmetrically substituted aromatic rings. The experimental antifertility
and antiviral agent gossypol contains such a system and hence can exist in two enantiomeric
forms (Fig. 11).
Finally, compounds that assume a helical shape, even a partial helix, may have left- and right-
handed orientations and thus are chiral. The most notable examples of such chirality are the
helices formed in polynucleotides (such as deoxyribonucleic acid) (DNA) and proteins.
N
N
CH
3
CH
3
COOH
CH
3
H
S
O
CH
3
F
COOH
H
3
C
H
S
O
CH
3
F
CH
3
H
Br
COOH
HOOC
HOOC
NO
2
NO
2
HOOC
HOOC
NO
2
NO
2
Cl
O
2
N
HOOC
NO
2
COOH
A B
C
Principles of Drug Action I, Winter 2005
6
However, drugs such as the benzodiazepine anxiolytics also exist as distinct helical forms (Fig.
12). Although these stereoisomers readily interconvert, only one helical form possesses affinity
for the benzodiazepine receptor.
Figure 11. The chiral biaryl system gossypol.
Figure 12. Helical benzodiazepine isomers
IV. Definitions and Nomenclature
As discussed in the preceding section, each member of an enantiomeric pair rotates the plane of
polarized light to the same degree, but in opposite directions (dextrorotatory and levorotatory).
However, the amount of optical rotation is not constant for an individual enantiomer but rather is
dependent on the solvent, concentration, temperature, the wavelength of light used, and the path
length of the sample cell employed to determine the rotation. Thus, meaningful optical rotation
comparisons for chiral compounds are only possible when optical activities are determined under
specified conditions. Such conditions are defined as specific rotations [or] and are expressed for
solutions and neat liquids in Eqs. (1) and (2), respectively.
(1) []t, = 100/(l x c)
(2) []t, = /(l X d)
where [] = measured rotation, t = temperature, = wavelength, c = concentration, d = density
and l = length
CH
3
OH
CHO
OH
HO
CH
3
CH
3
OH
CH
3
CH
3
CH
3
OH
OH
CHO
N
N
CH
3
O
Y
X
N
N
H
3
C
O
Y
X
Principles of Drug Action I, Winter 2005
7
Specific rotation data may assist in the identification of a specific enantiomer, or may be used to
determine the optical purity (enantiomeric purity) of a mixture of enantiomers. Optical purity is
defined as the percent excess of one enantiomer over another in a mixture and is expressed in Eq.
(3).
(3) Optical purity = ([d] - [l])/([d] + [l]) x (
obs
)/
o
Based on Eq. (3), a mixture consisting of equal amounts of each enantiomer would have no net
optical rotation; the optical rotation of one enantiomer is cancelled by the rotation of the other
enantiomer. Such a mixture is referred to as a racemic mixture or racemate.
Other terms commonly applied in discussions of optically active compounds include resolution
and racemization. Resolution describes the processes whereby a racemic mixture is separated
(resolved) into its component enantiomers. Racemization refers to the conversion of either
enantiomer into equal parts (racemic mixture) of both enantiomers.
Figure 13. Alanine and D/L configurational assignments
Over the years, several nomenclature systems have been developed to characterize the
relationship between enantiomers. The system based on optical activity and the classification of
enantiomers as dextrorotatory (d or (+)) or levorotatory (1 or (-)) already has been described.
However, this system of nomenclature is of limited applicability because the sign of rotation, (+)
or (-), does not predict the absolute configuration or the relative spatial arrangement of atoms in
the enantiomers. In an attempt to designate the precise configurations about carbon centers of
asymmetry, two nomenclature systems have been developed: the D/L system and the Cahn-
Ingold-Prelog R/S system. The application of these systems requires that chiral molecules be
oriented in a Fischer projection to reflect their tetrahedral geometry. In such a projection atoms
or groups on the horizontal axis are projecting out of the plane of the paper, whereas those on the
vertical axis are projecting behind the plane of the paper. When using the D/L system, the
projection is oriented in such a way that the main carbon chain is positioned vertically, with the
lowest numbered carbon positioned at the top. After the projection is aligned, the position of the
principal substituent relative to the carbon chain is identified. If it is to the left of the vertical
axis of the projection, the L-configuration is assigned. If the principal substituent is to the right
of the vertical axis, the D-configuration is assigned. The application of this system is
demonstrated in Fig.13 for the enantiomers of the amino acid alanine.
The D/L system was used widely in the past to assign the configurations of naturally occurring
compounds such as amino acids and sugars. Its application to unnatural structures, such as most
drugs, is limited because it can only be used when a compound contains a main carbon chain, and
when an unambiguous choice of the principal substituents can be made. The use of this system is
also complicated by the fact that the D- and L-absolute configurations are not related to the
COOH
CH
3
NH
2
H
COOH
CH
3
NH
2
H
COOH
CH
3
H
2
N H
COOH
CH
3
H
2
N H
D-Alanine
L-Alanine
Principles of Drug Action I, Winter 2005
8
directions that enantiomers rotate plane polarized light. For example, the D-enantiomer of
glyceraldehyde rotates plane-polarized light to the right (dextrorotatory), while U-glyceraldehyde
is levorotatory (Fig. 14). However, with the related enantiomers of glyceric acid, the reverse is
true. Thus structurally related compounds with the same absolute configuration do not
necessarily rotate the plane of polarized light in the same direction. This lack of correlation
between the D/L and d-(+)/1-(-) nomenclature systems remains a source of confusion and error in
the literature.
Figure 14. Configurations and optical rotations of enantiomers of glyceraldehyde and
glyceric acid
The lack of a relationship between the sign of optical rotation (d or 1) and absolute configuration
as designated by D/L, coupled with the uncertainties in assignments for primary substituents
inherent to the D/L nomenclature system, led to the development of an unambiguous system for
the designation of absolute configuration by Cahn, Ingold, and Prelog. In applying this method,
the compound is oriented in a Fischer projection and the four groups or atoms bound to an
asymmetric carbon are ranked by the following set of sequence rules:
1. Substituents are ranked (1, 2, 3, 4) by the atomic number of the atom directly joined to the
chiral carbon.
2. When two or more of the atoms connected to the chiral carbon are the same, the atomic
number of the next adjacent atom determines the priority. If two or more atoms connected to
the second atom are the same, the third atom determines the priority, etc.
3. All atoms except hydrogen are formally given a valence of 4. When the actual valence is less
than 4 (N, O), phantom atoms are assigned an atomic number of zero and therefore rank the
lowest.
4. A tritium atom has a higher priority than deuterium, which has a higher priority than
hydrogen. Similarly, any higher isotope has a higher priority than any lower one.
5. Atoms with double and triple bonds are counted as if they were connected by two or three
single bonds. Hence a C==C is regarded as a carbon bound to two carbons; and a C==O is
regarded as a carbon bound to two oxygens,
Once the four groups bound to the chiral carbon are ranked, the compound is oriented in such a
way that the lowest priority group (4) is projected away from the observer. Then, if the other
groups (1, 2, 3) are oriented by priority in a clockwise fashion, the molecule is designated as R
(rectus), and if counterclockwise, as S (sinister). These sequence rules are applied in the
assignment of the absolute configurations for the enantiomers of glyceraldehyde in Fig. 15.
According to the first rule, the highest priority substituent (1) is the hydroxy group (OH) and the
lowest priority group (4) is the hydrogen atom; since the atomic number of carbon is higher than
CHO
CH
2
OH
OH H
CHO
CH
2
OH
HO H
COOH
CH
2
OH
OH H
COOH
CH
2
OH
HO H
D-(+)-isomer L-(-)-isomer
D-(-)-isomer
L-(+)-isomer
GLyceralhdeyde
Glyceric Acid
Principles of Drug Action I, Winter 2005
9
that of hydrogen but lower than that of oxygen, the two carbon substituents (CHO and CH
2
OH)
are assigned intermediate priorities. To determine the priority relative to the two carbon
substituents, both the 2nd and 5th rules must be applied. The aldehyde carbon is part of a
carbonyl (C=O) moiety which, by rule 5, is equivalent to a carbon bound to two oxygen atoms.
The alcohol carbon (CH2OH) is bound to one oxygen and two hydrogens. The "two oxygens" of
the aldehyde take priority over the single oxygen of the alcohol moiety; the aldehyde is assigned
priority 2, and the alcohol priority 3. With all substituents ranked and the enantiomers oriented
in such a way that the lowest priority group (4) is projected away from the observer, the
configurations can be assigned. The enantiomer in which the substituents are oriented by priority
in a clockwise fashion, is designated as R, and the enantiomer in which the substituents are
oriented by priority in a counterclockwise direction is designated as S.
Figure 15. Cahn-Ingold-Prelog sequence rule
V. Compounds with Multiple Centers of Asymmetry
Many stereoisomeric drugs contain more than one asymmetrically substituted atom. For
example, ephedrine has two chiral centers, and the macrolide antibiotic eryth-romycin has 18
(Fig. 16). In such cases, a greater number of configurational isomers is possible;.the maximum
number possible is 2
n,
where n is the number of chiral atoms.
Figure 16. Compounds with multiple chiral centers
For compounds with two chiral centers, the maximum number of configurational isomers
possible is four, based on the 2
n
rule. Since each chiral center may have an R- or S-configuration,
the isomers possible include RR, SS, RS, and SR, as shown for ephedrine in Fig. 17. The RS and
SR isomers are nonsuperimposable mirror images and hence are enantiomers. The same
relationship exists for the RR and SS isomers. The relationship between each member of an
enantiomeric pair and each member of the other enantiomeric pair is diastereomeric; they are
H
CH
2
OH
OH
O
H
H
HO
O
H
HOCH
2
S-Glyceraldehyde R-Glyceraldehyde
1 2
3
4
4
1
2
3
Erythromycin
Ephedrine
HO
H
NHCH
3
CH
3
CH
3
OH
CH
3
OCH
3
O
O
CH
3 H
3
C
OH
OH
H
3
C
O
CH
3
OH
CH
3
O
O
C
2
H
5
O
O
HO
CH
3
N
CH
3
H
3
C
CH
3
Principles of Drug Action I, Winter 2005
10
non-superimposable nonmirror images. Thus the RS isomer is a diastereomer of the RR and SS
isomers, and the SS isomer is a diastereomer of the RS and SR isomers. In this case, the SR and
RS enantiomers are referred to as ephedrines, whereas the RR and SS enantiomers are called
pseudoephedrines. Diastereomers, unlike enantiomers, have' different physicochemical
properties. Therefore the ephedrines have different solubilities, melting points, etc., than the
pseudoephedrines.
Figure 17. Ephedrine and pseudoephedrine stereoisomers.
In addition to the R and S designations, compounds with two chiral centers may also be
identified by stereochemical nomenclature which describes the entire system. For example, the
erythro and threo nomenclature derived from carbohydrate chemistry may be employed to
describe the relative positions of similar groups on each chiral carbon. Thus, the ephedrines are
designated as erythro forms since the similar groups (OH and NHCH3) are on the same side of
the vertical axis of the Fischer projection, and the pseudoephedrines are designated as threo
forms since like groups are on opposite sites of the vertical axis of the projection (Fig. 17).
Figure 18. Ethambutol stereoisomers
It is important to realize that the 2
n
rule predicts only the maximum number of stereoisomers
possible in compounds with more than one center of chirality. For example, some compounds
with two asymmetrically substituted carbon atoms may have only three stereoisomeric forms.
This occurs when three of the substituents on one asymmetric carbon are the same as those on the
other asymmetric carbon, as shown for the antitubercular ethambutol (Fig. 18). In this case, one
CH
2
CH
3
CH
2
OH H
N H
CH
2
CH
2
N H
HOCH
2
H
CH
2
CH
3
CH
2
CH
3
HOCH
2
H
N H
CH
2
CH
2
N H
CH
2
OH H
CH
2
CH
3
CH
2
CH
3
H CH
2
OH
N H
CH
2
CH
2
N H
CH
2
OH H
CH
2
CH
3
(R,R)-Isomer
(S,S)-Isomer Meso Form
CH
3
NHCH
3
H
OH H
CH
3
CH
3
NH H
HO H
CH
3
CH
3
NH H
H OH
CH
3
NHCH
3
H
H HO
(1S, 2R) (1R, 2S) (1R, 2R)
(1S, 2S)
Ephedrines (Erythro forms)
Pseudoephedrines (Threo forms)
Principles of Drug Action I, Winter 2005
11
stereoisomer has a plane of symmetry even though two asymmetric atoms are present. Such an
isomer is referred to as a meso compound and is optically inactive. Therefore, when a plane of
symmetry is present in a compound with two centers of asymmetry, only three stereoisomeric
forms are possible.
A number of drugs possess three or more asymmetrically substituted carbon atoms. For example,
there are three chiral carbon atoms in the bicyclic ring system of the penicillin antibiotics (Fig.
19). Based on the 2
'
rule, eight stereoisomers are possible for the penicillins, but only the
naturally biosynthesized 3S,5R,6R en-antiomer displays significant antibacterial activity. Other
antibiotics, such as the aminoglycosides and the tetracyclines, contain multiple chiral centers; the
macrolide erythromycin has 18 chiral centers and thus there are 262,144 possible stereoisomers!
(Fig. 16). These antibiotics are produced by microbes from chiral endogenous substances and
thus a single stereoisomeric form is typically produced.
Figure 19. (3S, 5R, 6R)-Benzylpenicillin
5
6
7
1
2
3
4
N
S
N
H H
COOH
CH
3
CH
3
O
H
O
H